Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 May 1.
Published in final edited form as: Osteoarthritis Cartilage. 2013 Jan 31;21(5):710–720. doi: 10.1016/j.joca.2013.01.006

Radiofrequency (RF) Coil Impacts the Value and Reproducibility of Cartilage Spin-Spin (T2) Relaxation Time Measurements

BJ Dardzinski 1, E Schneider 2
PMCID: PMC3624070  NIHMSID: NIHMS441687  PMID: 23376528

Abstract

Introduction

T2 (spin-spin) relaxation time is frequently used for compositional assessment of articular cartilage. However little is known about the influence of MR system components on these measurements. The reproducibility and range of cartilage T2 values were evaluated using different extremity radiofrequency (RF) coils with potential differences in flip angle uniformity and SNR.

Method

Ten knees underwent 3 Tesla MR exams using RF coils with different signal-to-noise (SNR): quadrature transmit/receive (QTR); quadrature transmit/eight-channel phased-array receive (QT8PAR). Each knee was scanned twice per coil (4 exams total). T2 values were calculated for the central medial and lateral femoral (cMF, cLF) and medial and lateral tibial (MT, LT) cartilage.

Results

The flip angle varied across a central 40mm diameter region-of-interest of each coil by <1.5%. However SNR was significantly higher using QT8PAR than QTR (p<0.001). T2 values for cMF (50.7msec/45.9msec) and MT (48.2msec/41.6msec) were significantly longer with QT8PAR than QTR (p<0.05). T2 reproducibility was improved using QT8PAR for cMF and cLF (4.8%/5.8% and 4.1%/6.5%; p<0.001), similar for LT (3.8%/3.6%; p=1.0), and worse for MT (3.7%/3.3%; p<0.001). T2 varied spatially, with cLF having the longest (52.0msec) and the LT having the shortest (40.6msec) values. All deep cartilage had significantly longer, and less variable, T2 values using QT8PAR (higher SNR; p<0.03).

Conclusions

SNR varied spatially depending upon coil, but refocusing flip angle did not. With higher SNR, significantly longer T2 values were measured for deep (all plates) and global (MT, cMF) cartilage. T2 values varied by depth and plate, in agreement with prior studies.

Keywords: Magnetic Resonance, Signal-to-Noise, Cartilage, T2, Osteoarthritis Initiative, Radiofrequency Coil

Introduction

Magnetic Resonance (MR) imaging at field strengths above 1.5 Telsa (T) is desirable for quantitative articular cartilage morphology and composition measurements due to intrinsically higher available signal-to-noise ratio (SNR)[1,2]. While the spin-spin relaxation time, T2, is fairly constant at lower field strength[1,3–5,6] it decreases slightly above 1.5T (~10% at 3T and 10–20% at 4T)[1,3–5,7]. This is important as T2 relaxation times have been used to evaluate the biochemical status of articular cartilage[8] in both cross-sectional and longitudinal studies of osteoarthritis and cartilage repair[9–11]. Interpretation and comparison of T2 values is challenging due to the range of acquisition parameters and analysis methods used (Table 1)[3–18]. It is thus important to understand the variables, including MR system components that may influence T2 values.

Table 1.

Summary of T2 measurements at different magnetic field strengths in ‘normal’ cartilage in ‘healthy’ subjects. Mean T2 value +/− standard deviation is presented. MSME = multi-slice, multi echo; SE = spin echo. Global T2 values are presented; if available top-third (T), central-third (C), and deep (D) cartilage T2 values are provided.

Ref Field Strength Coil Type Sequence Patellar T2 (msec) MT T2 (msec) LT T2 (msec) MF T2 (msec) LF T2 (msec) Other T2 (msec)
Maier [12] 1.5T 3” receive Single echo, single slice SE 20.2 to 40.7
Maier [12] 1.5T 3” receive coil MSME SE 23.3 to 36.9
Dunn [44] 1.5T 4 channel PA TR Dual echo SE 32.1 ± 1.4 34.9 ± 1.8 34.9 ± 1.0 35.0 ± 1.1
Ghosh [45] 1.5T 4 channel PA TR Dual echo SE 32.1 ± 2.0 tibial 31.1 ± 2.4 femoral 35.1 ± 1.5
Liess [34] 1.5T 8cm circular receive surface coil Fat- suppressed MSME SE 23.7 ± 0.6
Glaser [46] 1.5T Quadrature MSME SE 32.4 ± 2.1; 36.3 ± 3.2 (T); 32.2 ± 2.4 (C); 28.8 ± 2.1 (D)
Klosterman [5] 1.5T Quadrature MSME SE 50.1 ± 1.5 (T); 39.7 ± 0.9 (C); 43.0 ± 1.2 (D)
Klosterman [5] 3T 1 channel TR MSME SE 52.1 ± 1.6 (T); 40.3 ± 1.2 (C); 45.2 ± 1.6 (D)
Gold [4] 1.5T 4 channel TR PA T2 prep 42.1 ± 7.05 muscle 35.3 ± 3.85; marrow 165 ± 4.96
Gold [47] 3T Quadrature T2 prep 36.9 ± 3.81 muscle 31.7 ± 1.90; marrow 133 ± 6.14
Joseph [9] 3T Quadrature MSME SE 29.5 ± 1.8 36.8 ±2.2
Mosher [10] 3T ≥4 channel PA receive MSME SE patellar 42.4 ± 2.8 (T); 37.9 ± 2.1 (C); 35.2 ± 1.9 (D) tibial 47.8 ± 1.6 (T); 43.8 ± 1.8 (C); 41.5 ± 2.0 (D) femoral 45.6 ± 1.9 (T); 40.6 ± 1.7 (C); 38.9 ± 2.1 (D)
Welsch [35] 3T 8 channel PA receive MSME SE 54.2 ± 6.9
Welsch [35] 3T 8 channel PA receive MSME SE 54.6 ± 7.2
Watanabe [13] 3T Quadrature MSME SE 37.1 ± 1.9; 34.9 ± 1.0 (T); 24.2 ± 0.8 (D)
Welsch [36] 7T Quadrature MSME SE 54.6 ± 13.0; 58.7 ± 13.1 (T); 50.4 ± 13.5 (D)
Welsch [36] 7T Quadrature MSME SE tibial 43.6 ± 8.5; 48.1 ± 9.2 (T); 39.1 ± 9.6 (D) femoral 56.3 ± 15.2; 58.3 ± 14.4 (T); 54.2 ± 17.5 (D)
Raya [37] 7T 28 channel PA receive Fat- suppressed MSME SE 22.9 ± 4.2; 24.8 ± 4.4 (T); 20.9 ± 4.3 (D)
Open in a new tab

The Osteoarthritis Initiative (OAI) opted to use 3T MR systems for cartilage morphometry and T2 relaxation time measurements [19, 21] on 4,794 men and women ages 45–79 who either have, or are at increased risk of developing, knee OA. These subjects were evaluated annually over 9 years with radiography and MR, along with biochemical, genetic and clinical assessments of disease activity. From baseline through the 72 month knee MR exams, the OAI used the same radiofrequency (RF) coil; for the 96 month MR exams, a new coil was used. For this reason, we investigated the impact of the coil on the value and reproducibility of cartilage T2 values by comparing measurements made at 3T using two extremity coils with different signal-to-noise (SNR) and signal reception: a quadrature transmit/receive (QTR: 0–72mos; USA Instruments, Aurora, OH, USA) and a quadrature transmit/eight-channel phased-array receive (QT8PAR: 96mos; InVivo Corp., Orlando, FL, USA)[20–22]. Quadrature-receive (QR) coils have fairly uniform SNR across the entire imaging field-of-view (FOV) although SNR drop-off toward the edges may be found. Phased-array receive (PAR) coils have higher SNR at the edges and similar SNR at the center compared to cylindrical quadrature coils. As a result, in a PAR coil, the SNR will vary across the knee and may contribute measurement variation as a function of cartilage plate.

We hypothesized that cartilage T2 relaxation time measurement reproducibility can be improved by using the PAR coil (QT8PAR) compared to a quadrature coil (QTR). And, secondarily, that T2 times obtained with the QT8PAR coil are consistent with those obtained with the QTR coil.

Method

Study Participants

The study was performed at two centers (Ohio State University and Memorial Hospital of Rhode Island) as part of an OAI pilot study[20]. The study protocol, amendments, and informed-consent documentation were reviewed and approved by the local institutional review boards.

Ten adult subjects (three men, seven women; five healthy, five with a clinical diagnosis of OA) underwent test-retest MR imaging of either their left or right knee, and two subjects (one healthy, one with OA) underwent imaging of both their right and left knees. In total, 12 knees were examined four times each. All participants were recruited for other OAI pilot MR studies[20]. Seven of the 10 subjects also participated in the OAI[23] and underwent bilateral fixed-flexion posterior-anterior (P/A) radiography[24].

MR Acquisition

Test-retest images were acquired on 3T MR systems (Siemens Magnetom Trio, Erlangen, Germany) using both QTR and QT8PAR knee coils. Imaging was performed as in the OAI[19,20,23] including: double-oblique coronal 3D Fast Low Angle Shot with water excitation (FLASH); sagittal 3D Dual Echo in the Steady State with water excitation (DESS); and a sagittal multi-slice, multi-echo spin echo (MSME-SE) acquisition for T2 relaxation time measurement. The MSME-SE acquisition (Figure 1) used a 120mm FOV, 3mm slice thickness, with in-plane spatial resolution 0.31mmx0.45mm, one average, 2700msec repetition time, 7 echoes (10, 20, 30, 40, 50, 60, 70msec) and was prescribed sagittal to the joint, co-planar with the DESS acquisition and orthogonal to the FLASH.

Figure 1.

Figure 1

Figure 1

Open in a new tab

Source multi-slice multi-echo spin echo (MSME-SE) images through the mid-point of the medial knee acquired using the quadrature transmit-receive knee (QTR) coil. Multiple contrast acquisitions having progressively longer TEs are combined to calculate T2 maps of the articular cartilage and adjacent tissues. These seven images illustrate how changing the TE affects the relative signal and relative contrast among the different tissues in the knee.

The coils were positioned 60mm right/left off the magnet isocenter. Subjects were positioned feet first and supine, with the inferior end of the patella located at coil isocenter[19,20,23]. For the QTR coil, the knee was slightly flexed (about 10°), with a cushion placed beneath the knee, and the heel positioned directly on the table. The knee angle on the QT8PAR coil was fixed, also at about 10°. With both coils the foot was secured in a vertical position with the great toe pointed straight up. To reduce misalignment between exams, the knee and foot position as well as the prescription was standardized. The DESS and MSME exams utilized identical angulation and center position[19,20,23].

The OAI knee phantom has two compartments, an outer cylinder with outer diameter and length 125mm and 128mm and an inner sphere with 57mm diameter. Each compartment contains a different concentration Magnevist (Schering AG, Germany) solution (sphere 10mM; cylinder 3.33mM) corresponding to the approximate T2 values of the deep and top layers of normal cartilage (18msec and 50msec, respectively)[50]. The OAI knee phantom was evaluated four times in the left and four times in the right coil position both for T2 value as well as for transmit uniformity using both RF coils (16 total exams).

Each subject underwent four MR exams. On one day, a test-retest examination was performed using one of the coils. Between the two exams the participant was removed from the magnet and allowed to walk for about 10min. On another day, within 1 month of the first MR exam, the same acquisitions were repeated using the other coil. The order of coil use was randomized. All MR images were reviewed for image quality by the MR technologist and were immediately reacquired if unacceptable (orientation, incomplete anatomical coverage, motion, artifact, etc.).

T2 Analysis and Region Selection

Since our purpose was to evaluate the system variables that affect cartilage T2, we eliminated knee compartments with visible signs of damaged cartilage. This approach improved the likelihood that a single exponential fit can assess the biological status of cartilage[16, 25] using clinical acquisitions. We focused on knees with cartilage having ‘normal’ appearance and 3 or more pixels (≥1mm) across the cartilage thickness. Visually intact surfaces on the DESS, FLASH and MSME-SE images without extreme thinning of cartilage were classified as ‘normal,’ although knees were not eliminated for having meniscal degeneration or tears and/or posterior or anterior cruciate ligament tears. Knees with one ‘damaged’ compartment (usually the patello-femoral joint) were eligible for analysis of the other two compartments; those knees having two or more ‘damaged’ compartments were ineligible for analysis.

T2 relaxation times were computed pixel-by-pixel from the MSME series using custom software (IDL, Exelis Visual Information Solutions, Boulder, CO). First, a threshold was manually determined (~10% of maximum signal intensity) and applied to remove the background noise. Next, a linear (2-point) fit to the log of the signal decay from the last 6 (of 7) echo images (Figure 1) was performed. The goodness of fit was evaluated using a residual map (T2 fit errors on a pixel-by-pixel basis) (Figure 2). A better fit has residual values closer to zero (darker gray scale and black pixels). Long T2 valued tissues, such as fluid, are expected to have poorer fits due to the limited range of echo times. Likewise regions of damaged cartilage might also be poorly represented by a single exponential fit[16,25].

Figure 2.

Figure 2

Figure 2

Open in a new tab

Example of T2 and Mo fits and the respective residual (error) maps from the QTR coil (A) T2 map, (C) T2 error map, (E) Mo map and (G) Mo error map, and the QT8PAR coil (B) T2 map, (D) T2 error map, (F) Mo map and (H) Mo error map. The QTR residual maps (C and G) have increased noise levels (brighter gray scale and white pixels) and more uniform noise. The QT8PAR coil residual maps (D and H) have non-uniform noise, but demonstrate a better fit with residual values closer to zero (darker gray scale and black pixels). Synovial fluid has a poorer fit than cartilage due to the acquisition parameters (white near menisci). Gray scale indicates residual errors from 0 to 100%.

To limit the center-to-edge variability of SNR caused by a PA receive coil[21,22], we focused on the central three slices of the lateral and medial femoral-tibial cartilage that were not covered by the meniscus (anterior-posterior). This provided four regions-of-interest (ROIs), one each on the medial and lateral tibial plateaus (MT, LT) and on the medial and lateral central femoral condyles (cMF, cLF). The first analyzed lateral slice (Figure 3A) was the fifth slice from the first lateral slice that contained bone. The first medial analyzed slice was the third slice from the first medial slice that contained bone. The femoral and tibial cartilage plates were analyzed using the same 3 slices for each side of the knee. The same anatomical slice locations were used for both coils.

Figure 3.

Figure 3

Open in a new tab

Images acquired using the quadrature transmit phased array receive (QT8PAR) coil with TE=10msec. The femoral and tibial cartilage plates were analyzed on the central three slices on the medial and lateral sides of the knee. A, B, C demonstrate the central three slices of the lateral joint compartment. Only the cartilage between the meniscus (between the red lines on D, E, F) was segmented and analyzed.

Four regions (MT, LT, cMF, cLF) in each knee were defined by manual cartilage segmentation from the T2 map and intercept images (Figures 3–4). A total of 192 regions were measured (2 coils, 12 knees, 2 exams, 4 regions per knee). The cartilage-subchondral bone interface was determined on the T2 map (Figure 4A) by the start of noise adjacent to the tibial cartilage (red arrows) and a combination of the start of the noise and the different grey scale values for the femoral cartilage (green arrows). The cartilage-joint interface was determined by the contrast changes on the intercept image (Figure 4B) to either fluid or adjacent cartilage (red lines). Only the cartilage between the meniscus (between the red lines on Figures 3D– 3F) was segmented and analyzed. Cartilage located underneath or above the meniscus was excluded.

Figure 4.

Figure 4

Figure 4

Figure 4

Open in a new tab

The cartilage-bone interface was determined using the calculated T2 image (A) by the adjacent noise caused by the cortical bone (note the femoral cortical bone contains a narrower noise band because the tibial cortical bone also contains the chemical shift artifact signal void). The interface was determined by the start of noise adjacent to the cartilage (see red arrows for tibial cartilage and green arrows for femoral cartilage). The cartilage-joint fluid or cartilage-cartilage interface was determined using the intercept image (B) by the contrast change from cartilage to either fluid or adjacent cartilage (see red lines at the cartilage-cartilage interface). Average T2 profiles normalized to 1.0 for thickness (0=subchondral bone, 1=articular surface) for the cMF cartilage plate using the (C) QTR and (D) QT8PAR coils. (E) Plot of T2 value as a function of cartilage depth, plate and RF coil.

After defining the regions, T2 relaxation profiles were generated by projecting the values on a line perpendicular to the subchondral bone[26–29]. An average T2 relaxation profile for each ROI was created; this average incorporated all the profiles from each of the three slices that were included in the analysis. All average profiles were normalized to 1.0 for thickness (0=subchondral bone, 1=articular surface; Figures 4C, 4D), and divided into 20 segments for analysis. This allowed variation of the cartilage T2 measurement to be determined as a function of normalized distance from the subchondral bone. The normalized, pooled profile was analyzed after excluding the first and last 4 points (1–4, 17–20) to minimize the effect of partial volume and chemical shift at the subchondral bone and the synovial fluid at the articular surface (Figure 4C and 4D). Next, the profile for each cartilage segment was divided into 3 sub-regions by averaging points 5–8 (deep-third), 9–12 (central-third), and 13–16 (upper-third). The average cartilage thickness was determined from the average relaxation profile for each plate.

Three additional regions were analyzed for quality control. These regions (Figure 5) were selected from the central medial slice and included a region in the tibial bone marrow, the infrapatellar fat pad, and the gastrocnemius. Fascial planes were avoided for the fat and muscle ROIs.

Figure 5.

Figure 5

Open in a new tab

Three additional regions of interest (ROIs) were analyzed for purposes of quality control. These ROIs were selected from the most central medial slice and included a region in the tibial bone marrow (A), the infrapatellar fat pad (B), and in the gastrocnemius (C). Fascial planes were avoided if possible for the fat and muscle ROIs. Signal-to-noise (SNR) was calculated based on the noise ROI located near the infra-patellar fat in (B).

Cartilage segmentation as well as T2 value and residual calculation was performed by one person in an unpaired manner (blinded to subject identification and coil pairing). Part of cartilage segmentation included the mean thickness measurement (ThC.me). Phantom analysis was performed in a similar manner, with a 40mm diameter ROI placed in the central compartment.

SNR Measurement

To understand the influence of SNR on T2 values, signal intensities (SI) in cartilage and bone marrow ROIs and noise levels of two ROIs were measured for both the medial and lateral sides (Figure 5). The bone marrow ROI spanned the entire proximal epiphysis. The cartilage ROI was as described. The noise ROI was outside the knee below the patella. To magnify the differences between the coils, the SI and noise were measured on images from the seventh echo (TE 70msec; Figure 1). The measurements from the test-retest acquisitions were used to compute the average SNR level for each coil.

Statistics and Computation of CVs

Bland-Altman plots of test-retest differences for T2 values measured using each coil were visually assessed for variance to mean relationships and out-of-bounds measures. Differences between the values measured using QTR and QT8PAR were tested for statistical significance using a paired Student’s t-test (null hypothesis of no difference in T2 between coils). Re-measurement reproducibility was analyzed by the root-mean-square coefficient-of-variation (RMS CV%) defined by Gluer et al.[30].

Results

Subjects

The 10 participants had mean age 52.2yrs (range: 45–73yrs) with mean body mass index (BMI) 28.2kg/m2 (range: 21.8–34.6kg/m2). Of the seven subjects who underwent bilateral knee radiography; five subjects had Kellgren-Lawrence grade (KLG) 1 knees, one knee had KLG 2, and one knee had KLG 3, using the screening, site radiograph interpretation. Both the KLG 2 and 3 knees had two or more compartments with MR evidence of cartilage abnormalities or thickness <3 pixels. Thus, ten femorotibial joints from 10 participants were evaluated.

SNR

Each cartilage region had 470-2237 pixels. There was a minimum of 300,000 pixels in each quality control region. The SNR (Table 2) was significantly higher using QT8PAR for all cartilage, muscle, infrapatellar fat, and marrow ROIs measured on the last echo images (TE=70msec).

Table 2.

Mean and standard deviation (+/− SD) for Signal-to-Noise Ratios (SNR) for bone marrow and cartilage of the medial and lateral tibia (MT and LT) from last echo (TE=70msec). P-values assess the difference between the mean SNR levels from the two RF coils (* indicates statistically significant difference between coils).

SNR MT Marrow * MT Cartilage * LT Marrow * LT Cartilage *
QTR Coil
Mean ± SD 33.3 ± 6.4 2.8 ± 0.7 32.4 ± 5.5 2.5 ± 0.3
QT8PAR Coil
Mean ± SD 42.6 ± 8.1 5.6 ± 0.6 47.4 ± 8.4 6.0 ± 0.6
P-value 0.03 <0.001 <0.001 <0.001
Open in a new tab

Phantoms

The mean+/− standard deviation T2 values were not statistically different (internal sphere) 18.62+/− 0.12msec and 18.79+/−0.25msec (p=0.33) and (external cylinder) 52.39+/−0.83msec and 51.59+/−0.78msec (p=0.06) using QTR and QT8PAR. The SNR over these ROIs were statistically different (sphere) 47.8+/−1.4 and 52.7+/−2.1 (p<0.001) and (cylinder) 111.1+/−4.4 and 139.0+/−13.8 (p<0.001) for QTR and QT8PAR respectively. The RF transmit uniformity (measured using a 720° pulse) was evaluated over a central 40mm diameter spherical ROI with <10° variation (<1.5%). For a series of eight 180° refocusing pulses, this corresponds to a 0.4% signal loss. No in vivo RF transmit maps were made; no correction for non-uniformity was attempted in the T2 analyses. The phantom T2 values were comparable between the two sites[50].

Cartilage T2 Relaxation Time

Bland-Altman plots of the test-retest T2 values (not shown) were unremarkable. The T2 relaxation times and measurement reproducibility are shown in Table 3 (Figure 4E). The global T2 values for cMF and MT as well as muscle were significantly longer with QT8PAR (p<0.0004). Due to the small number of knees, there was no significant difference in global T2 value between the coils for cLF (p=0.06); LT (p=1.0) and bone marrow (p=0.77) global T2 values were equivalent. The T2 residual plots (Figure 2) indicate a better single exponential fit was achieved (smaller error) using images acquired with QT8PAR (visual assessment).

Table 3.

Mean and standard deviation (+/− SD) for T2 relaxation times and root-mean-square coefficients-of-variation (RMS CV%). (A) Global T2 values for all ROIs are presented. (B) T2 values as a function of depth (deep-, central-, top-third) for all cartilage plates are presented. P-values are shown for differences between the mean T2 relaxation times from the two RF coils (* indicates statistically significant difference between coils).

Table 3A
T2 cLF Cartilage cMF Cartilage * LT Cartilage MT Cartilage * MT Marrow Fat Muscle *
QTR Coil
Mean ± SD (msec) 49.2 ± 5.2 45.9 ± 3.8 40.6 ± 3.5 41.6 ± 3.0 106.4 ± 2.6 97.0 ± 6.5 37.9 ± 2.1
RMS CV (%) 6.5 5.8 3.6 3.3 1.0 4.4 2.9
QT8PAR Coil
Mean ± SD (msec) 52.0 ± 3.8 50.7 ± 3.9 40.6 ± 3.8 48.2 ± 3.0 106.1 ± 3.8 94.4 ± 5.0 40.7 ± 2.4
RMS CV (%) 4.1 4.8 3.8 3.7 1.7 3.6 2.9
P-value 0.06 0.0003 1.0 0.0001 0.77 0.16 0.0004
Table 3B
T2 Deep-Third cLF Cartilage * cMF Cartilage * LT Cartilage * MT Cartilage *
QTR Coil
Mean ± SD (msec) 45.3 ± 6.3 40.3 ± 4.5 37.9 ± 4.7 37.9 ± 4.8
RMS CV (%) 12.4 5.6 6.2 4.0
QT8PAR Coil
Mean ± SD (msec) 49.9 ± 6.5 47.0 ± 4.7 45.1 ± 4.5 45.6 ± 4.2
RMS CV (%) 4.0 5.2 4.4 5.0
P-value 0.03 0.0001 0.0001 0.0001
T2 Central-Third cLF Cartilage cMF Cartilage * LT Cartilage * MT Cartilage *
QTR Coil
Mean ± SD (msec) 47.8 ± 6.2 42.4 ± 4.2 39.4 ± 3.7 40.1 ± 4.2
RMS CV (%) 9.4 4.9 5.8 8.3
QT8PAR Coil
Mean ± SD (msec) 50.4 ± 4.2 46.8 ± 4.5 44.6 ± 4.1 45.2 ± 3.7
RMS CV (%) 3.9 7.0 5.2 3.3
P-value 0.13 0.003 0.0002 0.0002
T2 Top-Third cLF Cartilage cMF Cartilage * LT Cartilage * MT Cartilage
QTR Coil
Mean ± SD (msec) 51.2 ± 5.8 48.3 ± 3.7 44.0 ± 3.6 47.6 ± 6.5
RMS CV (%) 5.5 5.2 4.7 9.8
QT8PAR Coil
Mean ± SD (msec) 52.8 ± 2.9 51.6 ± 4.2 47.1 ± 3.8 49.9 ± 3.8
RMS CV (%) 4.6 6.2 5.5 3.6
P-value 0.28 0.01 0.01 0.18
Open in a new tab

The T2 reproducibility was better for cLF, cMF, and infrapatellar fat using QT8PAR. The T2 reproducibility was slightly better for LT, MT, and marrow using QTR. For muscle regions, the RMS CV% was the same in both coils (Table 3A).

When the cartilage ROIs were divided into 3 depths (lower-third, central, upper-third), the T2 value increased from the subchondral bone to the articular surface for all cartilage plates and for both coils (Table 3B and Figure 4E). For the deep layer, T2 measurements made using QT8PAR were significantly longer (p<0.026) for all cartilage plates (cLF, cMF, LT, MT). The RMS CV% for the deep layer was smaller using QT8PAR for all cartilage plates except MT. For cartilage in the central-third, T2 measurements made using QT8PAR were significantly longer for cMF, LT and MT (p<0.003). The reproducibility for cartilage T2 in the central-third was smaller for cLF and MT using QT8PAR and for cMF and LT using QTR. For cartilage in the top layer, only cMF and LT had significantly longer T2 times using QT8PAR (p=0.012 and p=0.012). The reproducibility trend was the same as for the central cartilage.

No differences between sites were noted for any analyses.

Cartilage Thickness

The mean cartilage thickness was 1.6±0.3mm, 2.1±0.3mm, and 1.8±0.3mm for LT, MT and weight-bearing femoral cartilage (cLF and cMF combined) respectively. There were no statistically significant differences between thickness measurements from QT8PAR and QTR (p>0.2; data not shown).

Discussion

MR imaging of cartilage is challenging because it is quite thin and has a curved surface; the longitudinal quantification of its biochemical changes utilizing relaxation time measurements thus places increased demands on MR imaging practice and technology. We explored the impact of two extremity coils with different SNR characteristics on T2 measurements of knee cartilage at 3T. Ten middle-aged knees with relatively healthy cartilage were analyzed after test-retest imaging with both a quadrature and a phased array knee coil (four measurements). All femorotibial compartments included in the analysis did not have any MR visible cartilage defects or significant thinning.

A key factor that might alter T2 value is the RF coil transmit uniformity. Transmit uniformity is influenced by coil design (both had quadrature transmit), but also by patient electrical loading of the coil. In a phantom, no difference in a central 40mm diameter spherical ROI was found in transmit flip angle. No significant difference in phantom T2 value was found for this same region. In volunteers with visibly healthy cartilage, the cMF and MT cartilage and muscle T2 values were significantly (p<0.003) longer using QT8PAR. Deep cartilage had the largest change in T2 value (4.6–7.2msec). Significantly (p<0.03) higher SNR for all cartilage and marrow ROIs were found with QT8PAR. The impact of increased SNR was previously found[20] to result in measurement of increased cartilage volume (VC) with QT8PAR. Here, higher SNR levels did not significantly change cartilage thickness, but were found to increase T2 values (Table 3A), however the extent and significance of the increase was not consistent for all cartilage plates and depths. Smaller fit residuals were found (Figure 2) and smaller error bars were obtained on the profile plots using QT8PAR, indicating better fits were obtained with higher SNR.

Lower T2 RMS CV% were found for MT, LT, and marrow using QT8PAR. Overall we measured 3.7–11.1% RMS CV% with QTR and 3.3–6.5% with QT8PAR, lower than most prior 3T reproducibility measurements (Table 4) and most similar to that of Stahl et al.[11] using one MR system and better than achieved by Mosher et al.[10] using 5 MR systems. In our study, the T2 variability was limited to patient positioning, SNR, RF coil loading, MR system variability, reproducibility of cartilage segmentation and ROI selection.

Table 4.

Summary of T2 test-retest reproducibility at different magnetic field strengths in ‘normal’ cartilage in ‘healthy’ subjects. RMS CV% = root-mean-square coefficient of variation; MSME = multi-slice, multi echo; SE = spin echo. Global T2 values are presented; if available upper layer (T), central (C), and deep (D) cartilage T2 values are provided, some groups divide the cartilage thickness into two layers (T, D) and other divide into three layers (T, C, D).

Ref Cartilage Plate Field Strength Coil Type Sequence RMS CV%
Ghosh [45] Femoral 1.5T 4 channel PA TR Dual SE 1.5%
Ghosh [45] Tibial 1.5T 4 channel PA TR Dual SE 2.0%
Ghosh [45] Patellar 1.5T 4 channel PA TR Dual SE 4.4%
Liess [34] Patella 1.5T 8cm circularly polarized receive surface coil Fat- suppressed MSME SE 1.7% (average CV%)
Glaser [46] Patella 1.5T Quadrature MSME SE 3.2%
3.9% (D)
3.9% (C)
4.7% (T)
Welsch [35] Talar Trochlea 3T 8 channel PA receive MSME SE 3.2%
Welsch [35] Inferior Tibia 3T 8 channel PA receive MSME SE 4.7%
Mosher [28] Femoral tibial 3T linear MSME SE 10 – 15%
Mosher [33] Femoral tibial 3T linear MSME SE 1 – 3%
Mosher [10] Medial Femur 3T ≥4 channel PA receive MSME SE 9.4%
8.6% (D)
6.2% (C)
5.9% (T)
Mosher [10] Lateral Femur 3T ≥4 channel PA receive MSME SE 10.9%
7.6% (D)
6.3% (C)
6.0% (T)
Mosher [10] Medial Tibia 3T ≥4 channel PA receive MSME SE 9.2%
6.0% (D)
5.1% (C)
4.9% (T)
Mosher [10] Lateral Tibia 3T ≥4 channel PA receive MSME SE 8.1%
6.8% (D)
6.0% (C)
4.6% (T)
Mosher [10] Patella 3T ≥4 channel PA receive MSME SE 11.0%
7.3% (D)
6.3% (C)
7.3% (T)
Raya [37] Patella 7T 28 channel PA receive; linear transmit Fat- suppressed MSME SE 5.9%
5.9% (D)
5.8% (T)
Welsch [36] Femoral Tibial 7T Quadrature MSME SE 7.1%
6.5% (D)
7.7% (T)
Open in a new tab

A summary of T2 test-retest reproducibility is presented in Table 4 for ‘normal’ knee cartilage[5,4,9–13,28,33–37,44–47], most studies had a small number of subjects and/or analyzed only the patellar cartilage. Significantly intra- and inter-session reproducibility improvements were achieved by using a positioning device[28,33]. Comparison between T2 values of osteoarthritic patients and age/gender controls at 3T were performed by two groups: Stahl et al.[11] used a dual-echo fast-spin-echo T2 measurement with relatively low in-plane spatial resolution (0.625×0.625×3mm3), had <3% CV% reproducibility, and found systematically longer T2 values for all cartilage plates in the OA group, but reached statistical significance only in pLF (8 OA and 10 control subjects). The ACRIN multi-center trial[10] found improved reproducibility as the knee cartilage quality and quantity diminished. Based on four test-retest exams in 50 subjects, they also found systematic increasing T2 values for all cartilage depths and plates between normal, mild and moderate OA; however significance was only reached between the normal and moderate groups at all depths for cLF, cMF, and patellar cartilage, and between mild and moderate groups for all depths of the patella and for cMF and cLF deep- and central-layers.

In addition to differing sensitivities to SNR levels, the four cartilage ROIs also had significantly different global T2 values: varying from 40.6msec (LT) to 49.2msec (cLF) with QTR and 40.6msec (LT) to 52.0msec (cLF) with QT8PAR. Our cartilage T2 values (Table 3A) were similar to that measured[35] in the ankle (54msec) using identical hardware and MSME-SE acquisition, and were in general agreement with the ACRIN[10] results using several different manufacturers’ MR systems, acquisitions, and coils.

The depth trends are similar to prior manuscripts[13, 26–29] with T2 values increasing from deep-, central-, to top-third layers (subchondral bone to articular surface) for all cartilage plates and both coils (Table 3B and Figure 4E), although several recent publications found the central-third had the longest T2 values[5,10,36,37]. The deep cartilage had significantly longer T2 times with QT8PAR, however in the central-third, only cMF, LT and MT had longer T2 values, and in the top-third, only the cMF and LT had significantly longer T2 times. Since the depth variations in cartilage T2 are statistically different and also vary depending upon cartilage plate, a simple average over all cartilage plates or even a small region of interest cannot be used to represent the true T2 relaxation time. This is the case even when using a restricted subregion of a cartilage plate.

In addition to magnetic field strength and orientation[38–43], image acquisition and analysis methods[6,9,12–18] are known to impact the resultant cartilage T2 value. At 1.5T, Maier et al.[12] found the multi-echo multi-slice acquisition resulted in T2 values closest to the single-echo single-slice[6, 15]. Similarly Pai et al.[18] found cartilage T2 relaxation time varied depending on sequence, and was 28msec for SE compared to 45msec for fast-spin-echo (FSE) at 3T. Watanabe et al.[13] found the average T2 value measured by single-slice was longer than that measured by multi-slice. Our T2 values (Table 3A) were longer than measured by Gold et al.[4] (36.9msec patellar cartilage; 31.7msec muscle) using a 3T GE (General Electric Healthcare, Waukesha, WI) and QTR. In addition to changing absolute T2 times with acquisition sequence (single- versus multi-slice and SE versus FSE), the number of echoes and TE values play a role in measurement value and sensitivity to change. Some authors have used fat-suppressed imaging, others have used only a two echo acquisition. Another confounder of absolute T2 times is analysis method: a three-point time domain or a two-point natural log fit results in different values. Likewise, including or excluding the first data point (earliest TE) will result in different values depending upon the amount of stimulated echoes present[6,12,15]. Furthermore, Koff et al.[14] determined that fitting algorithms can produce different T2 values; non-linear calculations resulted in the shortest T2 values, linear fits were intermediate, and noise-weighted fits resulted in the longest values. The sensitivity of analysis algorithms to low SNR[17] was found to particularly impact the shorter deep cartilage values.

Limitations of this study include the small number of knees (n=10), increased variability because two sites were used for image acquisition, and only two measurements were made with each coil (Gluer et al.[30] advises a minimum of 14 knees measured 4 times). Other limitations include the visible health of the femoro-tibial cartilage: all patella-trochlea joints as well as femoral-tibial joints with cartilage damage were eliminated to enable a single exponential fit to be used for all plates. ROI selection and cartilage segmentation were both manual, which could introduce analysis errors. ROI variability may have contributed to the lack of statistically significant changes in the cLF central and upper as well as the MT upper layers of cartilage. ROIs were limited to small regions where magic angle effects and knee positioning would not dominate reproducibility [38–43] as well as to enable comparison to other studies. These results can be used as the lower limit of SNR-induced changes in T2 value over the knee. Use of the 2-point time-domain fit and its known sensitivity to noise is another limitation, although this methodology was selected to amplify the impact of noise. We did not perform in vivo B1 mapping on each subject and coil combination. The impact of subject loading on the refocusing flip angle is unknown, although with contemporary transmit extremity coils it is expected to be much less than when using transmit body coils. Although subject foot and leg position was well defined and slice orientation standardized, a positioner was not utilized and might have reduced measurement variability.

Our findings of variable T2 values resulting from use of different MR system components (extremity coil) leads to question “what is the true absolute T2 value?” The system influences include magnetic field strength, refocusing pulse flip angle (coil design and patient loading), and SNR. Low SNR results in underestimated values, particularly in the deep cartilage with shorter T2 values. We have shown that SNR impacts the different ranges of T2 values variably. In practice, SNR, spatial resolution and acquisition duration are tradeoffs. Use of high SNR images (with last echo image SNR above 2) and analysis methods insensitive to noise should enable reaching the “true” T2 value asymptotically. As with all quantitative imaging, measurement and interpretation of T2 values requires reproducible measurement and analysis approaches, including consistent MR system components. Our findings further imply that evaluations should be within subject, and should include internal landmarks and reference tissues that do not change with the disease process under investigation, if possible.

In conclusion, knee articular cartilage T2 values can vary with plate and coil, with the cLF condyle having the longest value and the LT plateau having the shortest value. Under conditions of higher SNR, significantly longer T2 values were measured; this was particularly evident for the deep cartilage layer as well as the cMF. This effect is the same order of magnitude as the impact of changing magnetic field strength.

Acknowledgments

The Osteoarthritis Initiative (OAI) and this pilot study are conducted and supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases in collaboration with the OAI Investigators and Consultants. The research reported in this article was supported in part by contracts N01-AR-2-2261, N01-AR-2-2262 and N01-AR-2-2258 from NIAMS.

We are grateful to the Ohio State University and Memorial Hospital of Rhode Island OAI study teams for recruitment of the study subjects and acquisition of the MR exams.

Role of the funding source: The funding source, NIAMS, had input into the study design and data collection methods, put did not participate in these activities. All data analysis and interpretation, manuscript writing and decision to submit for publication was the sole responsibility of the authors.

Abbreviations

3D

three-dimensional

ACRIN

American College of Radiology Imaging Network

cMF

central medial femur

cLF

central lateral femur

DESS

dual echo in the steady state

FLASH

fast low angle shot

FOV

field of view

KLG

Kellgren-Lawrence grade

LT

lateral tibia

MR

magnetic resonance

MSME-SE

multi-slice, multi-echo spin echo

MT

medial tibia

OAI

Osteoarthritis Initiative

PA

phased array

QTR

quadrature transmit / receive

QT8PAR

quadrature transmit / 8-channel phased array receive

RMS CV%

root mean square coefficient of variation

ROI

region of interest

SNR

signal to noise ratio

T

Tesla

T2

spin spin relaxation time, TE, echo time

TE

echo time

Footnotes

Author Contributions: All authors have made substantial contributions to all three sections below:

  1. the conception and design of the study (ES), or acquisition of data (ES), or analysis (BJD) and interpretation of data (BJD and ES)
  2. drafting the article or revising it critically for important intellectual content (BJD and ES)
  3. final approval of the version to be submitted (BJD and ES)

ES (schneie1@ccf.org) takes responsibility for the integrity of the work as a whole, from inception to finished article.

Conflict of interest: BJD and ES were supported in part by a subcontract from the OAI coordinating center contract (NIAMS/NIH N01-AR-2-2258) to perform this pilot study and analysis.

ES has a fee for service contract with NIAMS as the NIAMS OAI Technical Advisor.

BJD has no conflicts pertinent to this work.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Duewell SH, Ceckler TL, Ong K, Wen H, Jaffer FA, Chesnick SA, et al. Musculoskeletal MR imaging at 4 T and at 1. 5 T: comparison of relaxation times and image contrast. Radiology. 1995;196:551–555. doi: 10.1148/radiology.196.2.7617876. [DOI] [PubMed] [Google Scholar]
  • 2.Eckstein F, Mosher T, Hunter D. Imaging of knee osteoarthritis: data beyond the beauty. Curr Opin Rheumatol. 2007;19:435–443. doi: 10.1097/BOR.0b013e328248b4be. [DOI] [PubMed] [Google Scholar]
  • 3.Bottomley PA, Foster TH, Argersinger RE, Pfeifer LM. A review of normal tissue hydrogen NMR relaxation times and relaxation mechanisms from 1–100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Med Phys. 1984;11:425–448. doi: 10.1118/1.595535. [DOI] [PubMed] [Google Scholar]
  • 4.Gold GE, Han E, Stainsby J, Wright G, Brittain J, Beaulieu C. Musculoskeletal MRI at 3. 0 T: relaxation times and image contrast. AJR Am J Roentgenol. 2004;183:343–351. doi: 10.2214/ajr.183.2.1830343. [DOI] [PubMed] [Google Scholar]
  • 5.Klosterman LA, Schmithorst VJ, Tan S, Dardzinski BJ. T2 Measurement in Adult Patellar Cartilage at 1.5 and 3.0 Tesla. International Society of Magnetic Resonance in Medicine 7th Scientific Meeting & Exhibition; Philadelphia, PA. 1999. p. 1017. [Google Scholar]
  • 6.MacFall JR, Riederer SJ, Wang HZ. An analysis of noise propagation in computed T2, pseudodensity, and synthetic spin-echo images. Med Phys. 1986;13:285–292. doi: 10.1118/1.595956. [DOI] [PubMed] [Google Scholar]
  • 7.Stanisz GJ, Odrobina EE, Pun J, Escaravage M, Graham SJ, Bronskill MJ, et al. T1, T2 relaxation and magnetization transfer in tissue at 3T. Magn Reson Med. 2005;54:507–512. doi: 10.1002/mrm.20605. [DOI] [PubMed] [Google Scholar]
  • 8.Mosher TJ, Dardzinski BJ. Cartilage MRI T2 relaxation time mapping: overview and applications. Semin Musculoskelet Radiol. 2004;8:355–368. doi: 10.1055/s-2004-861764. [DOI] [PubMed] [Google Scholar]
  • 9.Joseph GB, Baum T, Carballido-Gamio J, Nardo L, Virayavanich W, Alizai H, et al. Texture analysis of cartilage T2 maps: individuals with risk factors for OA have higher and more heterogeneous knee cartilage MR T2 compared to normal controls - data from the osteoarthritis initiative. Arthritis Res Ther. 2011;13:R153. doi: 10.1186/ar3469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Mosher TJ, Zhang Z, Reddy R, Boudhar S, Milestone BN, Morrison WB, et al. Knee articular cartilage damage in osteoarthritis: analysis of MR image biomarker reproducibility in ACRIN-PA 4001 multicenter trial. Radiology. 2011;258:832–842. doi: 10.1148/radiol.10101174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Stahl R, Blumenkrantz G, Carballido-Gamio J, Zhao S, Munoz T, Hellio Le Graverand-Gastineau MP, et al. MRI-derived T2 relaxation times and cartilage morphometry of the tibio-femoral joint in subjects with and without osteoarthritis during a 1-year follow-up. Osteoarthritis Cartilage. 2007;15:1225–1234. doi: 10.1016/j.joca.2007.04.018. [DOI] [PubMed] [Google Scholar]
  • 12.Maier CF, Tan SG, Hariharan H, Potter HG. T2 quantitation of articular cartilage at 1. 5 T. J Magn Reson Imaging. 2003;17:358–364. doi: 10.1002/jmri.10263. [DOI] [PubMed] [Google Scholar]
  • 13.Watanabe A, Boesch C, Obata T, Anderson SE. Effect of multislice acquisition on T1 and T2 measurements of articular cartilage at 3T. J Magn Reson Imaging. 2007;26:109–117. doi: 10.1002/jmri.20962. [DOI] [PubMed] [Google Scholar]
  • 14.Koff MF, Amrami KK, Felmlee JP, Kaufman KR. Bias of cartilage T2 values related to method of calculation. Magn Reson Imaging. 2008;26:1236–1243. doi: 10.1016/j.mri.2008.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.MacFall JR, Wehrli FW, Breger RK, Johnson GA. Methodology for the measurement and analysis of relaxation times in proton imaging. Magn Reson Imaging. 1987;5:209–220. doi: 10.1016/0730-725x(87)90022-1. [DOI] [PubMed] [Google Scholar]
  • 16.Reiter DA, Lin PC, Fishbein KW, Spencer RG. Multicomponent T2 relaxation analysis in cartilage. Magn Reson Med. 2009;61:803–809. doi: 10.1002/mrm.21926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Raya JG, Dietrich O, Horng A, Weber J, Reiser MF, Glaser C. T2 measurement in articular cartilage: impact of the fitting method on accuracy and precision at low SNR. Magn Reson Med. 2010;63:181–193. doi: 10.1002/mrm.22178. [DOI] [PubMed] [Google Scholar]
  • 18.Pai A, Li X, Majumdar S. A comparative study at 3 T of sequence dependence of T2 quantitation in the knee. Magn Reson Imaging. 2008;26:1215–1220. doi: 10.1016/j.mri.2008.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Peterfy CG, Schneider E, Nevitt M. The osteoarthritis initiative: report on the design rationale for the magnetic resonance imaging protocol for the knee. Osteoarthritis Cartilage. 2008;16:1433–1441. doi: 10.1016/j.joca.2008.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Eckstein F, Kunz M, Hudelmaier M, Jackson R, Yu J, Eaton CB, et al. Impact of coil design on the contrast-to-noise ratio, precision, and consistency of quantitative cartilage morphometry at 3 Tesla: a pilot study for the osteoarthritis initiative. Magn Reson Med. 2007;57:448–454. doi: 10.1002/mrm.21146. [DOI] [PubMed] [Google Scholar]
  • 21.Hayes CE, Hattes N, Roemer PB. Volume imaging with MR phased arrays. Magn Reson Med. 1991;18:309–319. doi: 10.1002/mrm.1910180206. [DOI] [PubMed] [Google Scholar]
  • 22.Roemer PB, Edelstein WA, Hayes CE, Souza SP, Mueller OM. The NMR phased array. Magn Reson Med. 1990;16:192–225. doi: 10.1002/mrm.1910160203. [DOI] [PubMed] [Google Scholar]
  • 23.Osteoarthritis Initiative (OAI) Web Site: oai.epi-ucsf.org
  • 24.Kothari M, Guermazi A, von Ingersleben G, Miaux Y, Sieffert M, Block JE, et al. Fixed-flexion radiography of the knee provides reproducible joint space width measurements in osteoarthritis. Eur Radiol. 2004;14:1568–1573. doi: 10.1007/s00330-004-2312-6. [DOI] [PubMed] [Google Scholar]
  • 25.Reiter DA, Roque RA, Lin PC, Doty SB, Pleshko N, Spencer RG. Improved specificity of cartilage matrix evaluation using multiexponential transverse relaxation analysis applied to pathomimetically degraded cartilage. NMR Biomed. 2011;24:1286–1294. doi: 10.1002/nbm.1690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Dardzinski BJ, Mosher TJ, Li S, Van Slyke MA, Smith MB. Spatial variation of T2 in human articular cartilage. Radiology. 1997;205:546–550. doi: 10.1148/radiology.205.2.9356643. [DOI] [PubMed] [Google Scholar]
  • 27.Mosher TJ, Dardzinski BJ, Smith MB. Human Articular Cartilage: Influence of Aging and Early Symptomatic Degeneration on the Spatial Variation of T2-Preliminary Findings at 3 T1. Radiology. 2000;214:259–266. doi: 10.1148/radiology.214.1.r00ja15259. [DOI] [PubMed] [Google Scholar]
  • 28.Mosher TJ, Smith H, Collins C, Dardzinski BJ, Schmithorst VJ, Smith MB. Reproducibility of In Vivo Cartilage T2 Profiles: Implication for Longitudinal Studies. International Society of Magnetic Resonance in Medicine 9th Scientific Meeting & Exhibition; Glasgow, Scotland. 2001. p. 2097. [Google Scholar]
  • 29.Smith HE, Mosher TJ, Dardzinski BJ, Collins BG, Collins CM, Yang QX, et al. Spatial variation in cartilage T2 of the knee. J Magn Reson Imaging. 2001;14:50–55. doi: 10.1002/jmri.1150. [DOI] [PubMed] [Google Scholar]
  • 30.Gluer CC, Blake G, Lu Y, Blunt BA, Jergas M, Genant HK. Accurate assessment of precision errors: how to measure the reproducibility of bone densitometry techniques. Osteoporos Int. 1995;5:262–270. doi: 10.1007/BF01774016. [DOI] [PubMed] [Google Scholar]
  • 31.Nishii T, Tanaka H, Nakanishi K, Sugano N, Miki H, Yoshikawa H. Fat-Suppressed 3D Spoiled Gradient-Echo MRI and MDCT Arthrography of Articular Cartilage in Patients with Hip Dysplasia. Am J Roentgenol. 2005;185:379–385. doi: 10.2214/ajr.185.2.01850379. [DOI] [PubMed] [Google Scholar]
  • 32.Koff MF, Parratte S, Amrami KK, Kaufman KR. Examiner repeatability of patellar cartilage T2 values. Magn Reson Imaging. 2009;27:131–136. doi: 10.1016/j.mri.2008.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Mosher TJ, Liu Y, Smith MB. Improved Reliability of Cartilage T2 Measurements using a Leg Positioning Device. International Society of Magnetic Resonance in Medicine 7th Scientific Meeting & Exhibition; Kyoto, Japan. 2004. p. 816. [Google Scholar]
  • 34.Liess C, Lusse S, Karger N, Heller M, Gluer CC. Detection of changes in cartilage water content using MRI T2-mapping in vivo. Osteoarthritis Cartilage. 2002;10:907–913. doi: 10.1053/joca.2002.0847. [DOI] [PubMed] [Google Scholar]
  • 35.Welsch GH, Mamisch TC, Weber M, Horger W, Bohndorf K, Trattnig S. High-resolution morphological and biochemical imaging of articular cartilage of the ankle joint at 3. 0 T using a new dedicated phased array coil: in vivo reproducibility study. Skeletal Radiol. 2008;37:519–526. doi: 10.1007/s00256-008-0474-z. [DOI] [PubMed] [Google Scholar]
  • 36.Welsch GH, Mamisch TC, Hughes T, Zilkens C, Quirbach S, Scheffler K, et al. In vivo biochemical 7. 0 Tesla magnetic resonance: preliminary results of dGEMRIC, zonal T2, and T2* mapping of articular cartilage. Invest Radiol. 2008;43:619–626. doi: 10.1097/RLI.0b013e31817e9122. [DOI] [PubMed] [Google Scholar]
  • 37.Raya JG, Horng A, Dietrich O, Krasnokutsky S, Beltran LS, Storey P, et al. Articular cartilage: in vivo diffusion-tensor imaging. Radiology. 2012;262:550–559. doi: 10.1148/radiol.11110821. [DOI] [PubMed] [Google Scholar]
  • 38.Goodwin DW, Wadghiri YZ, Zhu H, Vinton CJ, Smith ED, Dunn JF. Macroscopic Structure of Articular Cartilage of the Tibial Plateau: Influence of a Characteristic Matrix Architecture on MRI Appearance. Am J Roentgenol. 2004;182:311–318. doi: 10.2214/ajr.182.2.1820311. [DOI] [PubMed] [Google Scholar]
  • 39.Goodwin DW, Dunn JF, Mosher TJ, Smith HE, Dardzinski BJ. MR Imaging and T2 Mapping of Femoral Cartilage. Am J Roentgenol. 2002;178:1568–1570. doi: 10.2214/ajr.178.6.1781568. [DOI] [PubMed] [Google Scholar]
  • 40.Xia Y, Moody JB, Alhadlaq H. Orientational dependence of T2 relaxation in articular cartilage: A microscopic MRI (microMRI) study. Magn Reson Med. 2002;48:460–469. doi: 10.1002/mrm.10216. [DOI] [PubMed] [Google Scholar]
  • 41.Mosher TJ, Smith H, Dardzinski BJ, Schmithorst VJ, Smith MB. MR Imaging and T2 Mapping of Femoral Cartilage: In Vivo Determination of the Magic Angle Effect. Am J Roentgenol. 2001;177:665–669. doi: 10.2214/ajr.177.3.1770665. [DOI] [PubMed] [Google Scholar]
  • 42.Mlynarik V, Mosher TJ, Smith HE, Dardzinski B. Magic Angle Effect in Articular Cartilage. Am J Roentgenol. 2002;178:1287–1288. doi: 10.2214/ajr.178.5.1781287. [DOI] [PubMed] [Google Scholar]
  • 43.Watanabe A, Boesch C, Siebenrock K, Obata T, Anderson SE. T2 mapping of hip articular cartilage in healthy volunteers at 3T: a study of topographic variation. J Magn Reson Imaging. 2007;26:165–171. doi: 10.1002/jmri.21014. [DOI] [PubMed] [Google Scholar]
  • 44.Dunn TC, Lu Y, Jin H, Ries MD, Majumdar S. T2 Relaxation Time of Cartilage at MR Imaging: Comparison with Severity of Knee Osteoarthritis. Radiology. 2004;232:592–598. doi: 10.1148/radiol.2322030976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Ghosh S, Beuf O, Ries M, Lane N, Majumdar S. Magnetic Resonance Imaging Based Evaluation of Articular Cartilage Degeneration. 47th Annual Meeting, Orthopaedic Research Society; San Francisco, CA. 2007. p. 0231. [Google Scholar]
  • 46.Glaser C, Mendlik T, Dinges J, Weber J, Stahl R, Trumm C, et al. Global and regional reproducibility of T2 relaxation time measurements in human patellar cartilage. Magn Reson Med. 2006;56:527–534. doi: 10.1002/mrm.21005. [DOI] [PubMed] [Google Scholar]
  • 47.Miller KL, Hargreaves BA, Gold GE, Pauly JM. Steady-state diffusion-weighted imaging of in vivo knee cartilage. Magn Reson Med. 2004;51:394–398. doi: 10.1002/mrm.10696. [DOI] [PubMed] [Google Scholar]
  • 48.Mosher TJ, Liu Y, Torok CM. Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running. Osteoarthritis Cartilage. 2010;18:358–364. doi: 10.1016/j.joca.2009.11.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Hannila I, Nieminen MT, Rauvala E, Tervonen O, Ojala R. Patellar Cartilage Lesions: Comparison of Magnetic Resonance Imaging and T2 Relaxation-Time Mapping. Acta Radiol. 2007;4:444–448. doi: 10.1080/02841850701280817. [DOI] [PubMed] [Google Scholar]
  • 50.Schneider E, NessAiver M, White D, et al. The Osteoarthritis Initiative (OAI) Magnetic Resonance Imaging Quality Assurance Methods and Results. Osteoarth Cart. 2008;16:994–1004. doi: 10.1016/j.joca.2008.02.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES